This application requests support for a 12 month period to complete the data analysis phase of the study "Determinants of Recovery Among Older Women". Data collection and data entry and editing have been completed; analysis of substantive issues are in progress. Funds are explicitly requested for partial support of the efforts of key study investigators and data analysis staff. The specific aims of this study have been to determine importance the degree, rate and course of recovery achieved after fracture of the hip. Spectra emphasis is placed on identification of the independent effects on recovery of depression following fracture, self rated health, personality and social supports. In addition, questions concerning the effects of the hip fracture and caregiving to the patient on the physical health status and emotional well- being of a non-patient, significant other will be addressed. Two hundred twenty-one white female hip fracture patients surgically treated for hip fracture in one of 17 Philadelphia area hospitals between August 1984 and January 1986 and 201 non- patients, significant others, were enrolled and followed for 12 months. After baseline interviews, subjects were reinterviewed, in-person, 2 and 6 months following surgery. Patients were interviewed at 12 months by telephone. Response rates for patients at two and six months respectively were 89.6 and 87.8 percent. Similar follow-up rates were achieved for non-patients. AT 12 months telephone Interviews (with patient, 80.5%; with proxy, 10.4%) were conducted. Of the remaining 221 original 7.7% were known to be dead with loss to follow-up limited to 1.4% (n=3). Primary analyses will focus on identification of factors that affect degree of recovery (defined interms of survival status, residence, walking ability, neuromuscular status, physical functioning, performance of self care and instrumental activities of daily living, presence of pain in the involved limb and subject assessment of satisfaction with recovery) at 6 and 12 months. Logistic and OLS regression procedures of ANOV techniques will be used for these analyses. Investigations of the course of recovery will employ causal models incorporating variables measured at different points in time. For these analyses we will rely heavily on the use of the LISRELVI procedures.